856295 - Document.readyState should be an enum

Status: NEW

(Core :: DOM: Core & HTML, defect, P5)

Description

[:Ms2ger (he/him; ⌚ UTC+1/+2)]

http://html5.org/tools/web-apps-tracker?from=7780&to=7781

Comment 1

[:Ms2ger (he/him; ⌚ UTC+1/+2)]

Hmm, it looks like we could return "uninitialized" in some cases. Henri, do you know if script can hit this?

Comment 2

[Henri Sivonen (:hsivonen)]

We are not supposed to return uninitialized in any script-detectable case, IIRC. What case do you have in mind?

Internally, we should keep a "not yet loading" state to detect bugs. Also, a "not yet loading into a document" state will be useful when moving readystate off nsDocument into a navigation life cycle object that the document points to as part of fixing bug 779959.

What's the benefit and motivation of the spec change?

Comment 3

[:Ms2ger (he/him; ⌚ UTC+1/+2)]

The benefit for us is that you can return an enum value and don't need to deal with strings in nsDocument.

Comment 4

[Sami Jaktholm]

Here's a one-liner that gives a script readyState of "uninitialized":
document.body.innerHTML = '<iframe src="/foo">'; document.querySelector("iframe").contentDocument.readyState

I think that bug 1171919 has hitting this case (readyState === "uninitialized") in a real document load setting (just a guess, don't know for sure).

Comment 5

[:Ms2ger (he/him; ⌚ UTC+1/+2)]

Reminder to ask Henri if he wants a bug for that when he's back.

Comment 6

[Henri Sivonen (:hsivonen)]

Thanks. Filed as bug 1191683.

Comment 7

[Jens Stutte [:jstutte]]

Bulk-downgrade of unassigned, >=5 years untouched DOM/Storage bugs' priority and severity.

If you have reason to believe this is wrong, please write a comment and ni :jstutte.